Predicting Effects of Operating Conditions on Biomass Fast Pyrolysis

Dec 2, 2016 - The numerical results showed that the conversion time increased as the ... found.26,27 Direct numerical simulation (DNS) is a promising...
0 downloads 0 Views 2MB Size
Subscriber access provided by University of Otago Library

Article

Predicting effects of operating conditions on biomass fast pyrolysis using particle-level simulation Yaoyu Pan, and Song-Charng Kong Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.6b02445 • Publication Date (Web): 02 Dec 2016 Downloaded from http://pubs.acs.org on December 6, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Predicting effects of operating conditions on biomass fast pyrolysis using particle-level simulation Yaoyu Pan, Song-Charng Kong* Department of Mechanical Engineering, Iowa State University, Ames, IA 50011, USA *Corresponding author. Email: [email protected]; Tel: +1 515-294-3244 Abstract Fast pyrolysis of biomass materials is an effective means to convert biomass into useful energy products. The conversion process can be significantly affected by the properties of the biomass particle and the operating conditions. To obtain a better understanding of this process, a direct numerical simulation method was proposed and used to conduct particle-scale simulations. In this study, the Lattice Boltzmann method was employed to solve the flow field and the intraparticle transport of heat and mass. A multi-step pyrolysis kinetics mechanism was used to describe the chemical reactions that convert solid biomass to gaseous and solid products. The predicted evolutions of center temperature and solid mass fraction agreed well with the experimental data. The validation demonstrated that the present model was capable of revealing the detailed conversion process of biomass fast pyrolysis at the particle scale. Parametric studies were conducted to characterize the effects of particle size, particle aspect ratio, inlet gas temperature, and reactor wall temperature on the conversion time and final product yields. The numerical results showed that the conversion time increased as the particle size increased and decreased as the inlet gas temperature and reactor wall temperature increased. When the particle size was decreased, more tar and syngas were produced while less char was generated. The same

1

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

trend of final product yields was also found when the inlet gas temperature and reactor wall temperature were increased. The results also indicated that the temperature gradients inside the particle can be neglected under certain particle size, i.e., equal to or less than 0.2 mm under the conditions studied. The heat flux from the reactor wall was found to be more significant to the fast pyrolysis process than the inlet gas temperature. Keywords: Biomass fast pyrolysis; Lattice Boltzmann method; Particle scale modeling

1

Introduction The increasing demand for energy and the importance of protecting the environment have

stimulated the utilization of alternative and sustainable energy sources to reduce the reliance on fossil fuels. As an abundant and renewable energy resource, biomass has low net carbon emissions compared to fossil fuels. Biomass has been widely used for heat, power and biofuel production by thermochemical and biochemical methods1. Fast pyrolysis is a promising thermochemical approach to produce various energy products from the low-energy density solid organic materials and has been an important topic in bioenergy research2,3. When exposed to an oxygen-free environment at high temperatures, the nonfood lignocellulosic biomass will be rapidly decomposed and form three primary products: condensable vapors (tar), non-condensable gases (syngas), and char (or biochar). The condensable vapors are further condensed to form biooil (or pyrolysis oil), an energy-intense liquid that can be further upgraded to transportation fuels4,5. Therefore, the understanding of the fundamental mechanisms of biomass fast pyrolysis is of practical importance but remains challenging6. The fast pyrolysis process of biomass particles in chemical reactors, such as bubbling fluidized-bed reactors and auger reactors, is extremely complex. The process involves multi-

2

ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

scale, multi-phase hydrodynamics, heat transfer, and chemical reactions. In addition to experimental study7,8, numerical simulation has become an effective tool to investigate the characteristics of this process9-17. Computational fluid dynamics (CFD) modeling, which employs conservation equations to describe the complex multiphase fluid flow and chemical reactions, has provided useful guidelines for optimizing reactor designs and operating conditions18-23. To evaluate the performance of a pyrolysis reactor, reactor-scale CFD modeling is conducted. However, the accuracy relies on the fidelity of the subgrid closures, such as intraparticle transport and gas-solid interactions. Descriptions of these detailed phenomena can have significant uncertainties since some of the submodels are derived empirically20,23,24. Because of the multi-scale feature of biomass particles and the complex multiphase flow in the reactor, it is critical to study the fast pyrolysis process of biomass at the particle scale12. Using detailed numerical simulations to derive appropriate subgrid closures can be a feasible strategy due to the difficulty of experimental study at the particle level. Various numerical models have been developed to describe the fast pyrolysis phenomena of a biomass particle9,10,13,15,25. However, to accurately predict the fast pyrolysis process of biomass at the particle scale, the intra-particle transport phenomena need to be resolved simultaneously with the surrounding gas flow. But only a few such studies in the literature were found26,27. Direct numerical simulation (DNS) is a promising approach to gain a detailed insight in biomass fast pyrolysis. In DNS, the Navier-Stokes equations are solved without the use of submodels, and the grid and time scales are small enough to capture the details of the flow. The Lattice Boltzmann method (LBM) is an effective numerical tool for studying multiphase fluid flow problems, particularly when dealing with the complex geometry and boundary28,29. The issue of complex boundaries will be encountered frequently in biomass fast pyrolysis, since the biomass

3

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

particle will change its shape rapidly and the shape is highly irregular, depending on the local conditions. This method was successfully applied to investigate the fluid flow and heat and mass transfer coupled with the surface chemical reactions in the porous media30,31. It has the potential to simulate the intra-particle transport phenomena of mass, heat, and species flows. The LBM can be further extended to model the gas-particle flow when coupled with discrete element method32, which makes it a viable approach to study the gas-solid interactions during biomass fast pyrolysis process. In the present study, a numerical framework was formulated and applied to simulate the biomass fast pyrolysis at the particle scale. The Lattice Boltzmann method was employed to solve the flow field and the intra-particle mass and heat transfer. A multi-step pyrolysis kinetics mechanism was used to describe the chemical reactions occurring inside the biomass particle during fast pyrolysis. The model was first validated by comparing the simulation results with the experimental data from the literature15,33. Previous numerical studies on the effect of operating conditions on the reactor performance was mainly based on reactor-scale CFD modeling, which used various empirical submodels18-23. This work will be focused on the effects of particle size, particle aspect ratio, inlet gas temperature, and reactor wall temperature on the biomass fast pyrolysis process.

2 2.1

Numerical methods Governing equations This study is focused on a stationary woody biomass particle exposed in a high-temperature

environment during fast pyrolysis. As the particle temperature increases, the virgin wood will be decomposed to tar, syngas, and char through a series of chemical reactions. The gaseous species

4

ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

(tar and syngas) will escape from the particle and merge into the surrounding gas. The hydrodynamics and heat transfer of the surrounding gas can be described by a set of conservation equations23, as given below. Assumptions, such as incompressible ideal gas and laminar flow, were applied, according to the experimental conditions [15].

∂ρg ∂t

+∇⋅ ( ρg ug ) = 0

∂(ρg ug ) ∂t

+∇⋅ (ρg ug ug ) = −∇p + µg ∇2ug

∂(ρg CPgTg ) ∂t

(1)

(2)

+∇⋅ (ρg CPgTg ug ) = λg ∇2Tg

(3)

To simulate the reaction kinetics of the decomposition of the particle, the wood pyrolysis model proposed by Park et al.15 was used in this study, as shown in Fig. 1. This model uses three competitive primary reactions to describe the conversion of virgin biomass to tar, syngas, and intermediate solid. The intermediate solid is further converted to char. The cracking reaction of tar further leads to the secondary syngas and char via two competitive reactions. All of the reaction rates are calculated using the first-order irreversible Arrhenius expression as

k = A exp( − E / RT ) . Kinetic parameters are listed in Table 115.

Fig. 1 Chemical reaction schemes of the wood pyrolysis model15

5

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 33

Table 1 Kinetic parameters15

Reaction

t

-1

Ai (s )

1.08×10

syn 10

4.38×10

is 9

3.75×10

c 6

1.38×10

c2 10

1.0×10

syn2 5

4.28×106

Ei (J/mol)

148000

152700

111700

161000

108000

108000

∆hi (kJ/kg)

80

80

80

-300

-42

-42

To describe the overall thermochemical conversion process of biomass particle, the pyrolysis kinetics is coupled with the conservation equations of intra-particle mass and heat transfer, as shown below15. Conservation of mass:

Virgin wood:

∂ρb = ωb = −(kt + ksyn + kis )ρb ∂t

Intermediate solid:

Char:

Tar:

(4)

∂ρis = ωis = kis ρb − kc ρis ∂t

(5)

∂ρc = ωc = kc ρis + kc2 ρt ∂t

(6)

∂(ερt ) + ∇⋅ ( ρt u f ) = ωt = kt ρb − (kc2 + ksyn2 ) ρt ∂t

Syngas:

∂(ερsyn ) ∂t

+ ∇ ⋅ ( ρsyn u f ) = ωsyn = ksyn ρb + ksyn 2 ρt

(7)

(8)

Conservation of energy:

(ρbCPw + ρisCPw + ρcCPc +ερt CPt +ερsynCPsyn )

∂Ts + (ρtCPt + ρsynCPsyn )∇⋅ (Ts uf ) = λs∇2Ts + Q ∂t

Q = − (kt ∆ht + ksyn ∆hsyn + kis ∆his )ρb − kc ∆hc ρis − (kc 2∆hc2 + ksyn2∆hsyn2 )ρt

6

ACS Paragon Plus Environment

(9) (10)

Page 7 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

In the above equations, ε is the porosity of biomass particle, defined as ε = 1 −

and

ρs (1 − ε w ) . ρs ρw

ρw are the total solid (sum of ρb , ρis and ρc ) and virgin biomass mass per unit volume. εw

is the initial porosity of biomass particle. To solve the intra-particle transport equations coupled with surrounding gas flow, it is assumed that the velocity of the gaseous species (tar and syngas) is equal to the velocity of surrounding gas at the particle surface ( u f = ug ). The gas flow in the porous biomass particle is calculated using the Darcy’s law. The external forces (i.e. gravity) are not considered in this simulation as the particle is fixed in the reactor. The surrounding gas is incompressible, ideal gas, and laminar flow. Particle shrinkage is not considered due to its negligible impact on the mass loss rate25. As shown in Eq. (9), local thermal equilibrium is assumed between the solid and gas phases15,25. Constant thermophysical properties of the components are used34, and thermal conductivity of the local solid mixture is linearly interpolated between the virgin biomass and the char formed. These thermophysical properties are listed in Table 2.

λ = (1 −η )λw + ηλc

(11)

The conversion extent of pyrolysis is defined as η = 1−

2.2

ρb + ρis . ρw

Lattice Boltzmann method In this study, the Lattice Boltzmann method (LBM) was used28,29. In the LBM, the

simulation domain is discretized into regular lattices, as shown in Fig. 2(a). The distribution functions of fluid properties are used on all lattices to recover the system’s hydrodynamics based

7

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 33

on the discrete form of the Boltzmann equations. By using the single-relaxation-time BhatnagarGross-Krook approximation35, the linearized Boltzmann equation for density is given by

gi ( x + ei ∆t , t + ∆t ) − gi ( x, t ) = −

1

τg

(gi ( x, t ) − gieq ( x, t ))

(12)

where x is the location of the lattice, t is the current time, and i is the discretized direction, ei is the discrete velocity along the i th direction, ∆t is the time step, eq

relaxation time related to the kinematic viscosity ν , and g i

τ g is the dimensionless

is the equilibrium density

distribution function. In this study, the D2Q9 model was adopted, in which the equilibrium density distribution function is defined as

gieq = ξi ρg (1+ 3

ei ⋅ u 9(ei ⋅ u)2 3u2 + − 2 ), c2 2c4 2c

i =0~8

where c is the lattice speed (defined as ∆x / ∆t , in which ∆x is the lattice spacing),

(13)

ξi

is the

weighting factor in i th direction. Macroscopic variables, such as density, velocity and pressure, can be recovered as

ρg = ∑ gi , i

ρg u = ∑ gi ei ,

p = cs2 ρg

(14)

i

where cs is the speed of sound. It has been proven that the Chapman-Enskog expansion of the Eq. (12) at low Mach number results in the incompressible Navier-Stokes equations as Eq. (1) and (2). More details of LBM can be found in previous works28,29.

8

ACS Paragon Plus Environment

Page 9 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

(a)

(b)

Fig. 2 (a) Schematic of domain discretization by LBM (b) description of particle boundary in lattice (corresponding to the small red circle in (a))

The evolution of the temperature distribution function has a similar form36,

Ti ( x + ei ∆t , t + ∆t ) − Ti ( x, t ) = −

where

τT

1

τT

(Ti ( x, t ) − Ti eq ( x, t )) + ξi

Q ρC p

is dimensionless relaxation time related to the thermal diffusivity

(15)

α

, which is

determined by the local lattice properties (fluid or solid), and Q is the source term due to N

chemical reactions ( Q = ∑ ∆hiωi , see Eq. (10)). The equilibrium temperature distribution i =1

function is formulated as follows37.

Ti eq = ξ i T (1 + 3

ei ⋅ u ), c2

i=0~8

(16)

The transport of the gaseous species (i.e., tar, syngas) is simulated by using LBM, while the solid species (i.e. biomass, char) remain at the fixed lattice with densities changing with time during reactions. In other words, the densities of the solid species are stored at the solid node while the gaseous species will propagate and diffuse among the solid node and the fluid node, as shown in Fig. 2(b). As shown in Fig. 2, the method used does not exactly capture the interface since a Cartesian grid is used.

9

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 33

The evolution of the density distribution function of the fluid phase is

f i ( x + ei ∆t , t + ∆t ) − f i ( x , t ) = −

1

τf

( f i ( x , t ) − f i eq ( x , t )) + ξiω f

where τ f is dimensionless relaxation time related to mass diffusion coefficient D and

(17)

ωf is the

source term due to chemical reactions (see Eq. (7) and Eq. (8) for tar and syngas). The equilibrium density distribution function is the same as Eq. (13).

fi eq = ξi ρ f (1+ 3

ei ⋅ u 9(ei ⋅ u)2 3u2 + − 2 ), c2 2c4 2c

i =0~8

(18)

Given these property distribution functions, the temperature T and species density ρf , can be obtained.

T = ∑Ti , i

ρ f = ∑ fi

(19)

i

The relationship between the relaxation time and the physical properties are given as follows. ν =

1 2 1 1 1 1 1 c (τ g − ) ∆ t , α = c 2 (τ T − ) ∆ t , D = c 2 (τ f − ) ∆ t 3 2 3 2 3 2

(20)

Boundary conditions at the inlet include the fully-developed parabolic velocity profile with a peak value Uin and a constant temperature Tin, as shown in Fig. 2(a). This is assumption is based on the premise that the present simulation domain represents a small space in a real reactor. At the outlet, constant pressure Pout is specified. The wall temperature Twall is assigned, and noslip boundary conditions are employed. The detailed treatments for addressing the unknown distribution functions on the boundary can be found in previous studies38-40. The computer code was written in C++ and run Dell Precision T5600 workstations. 2.3

Improved 2D model

10

ACS Paragon Plus Environment

Page 11 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

As described earlier, 2D LBM was used. For 2D simulation, the cylindrical particle is simplified as a circle, i.e., the cross section of the cylinder. But the spherical particle needs to be treated differently. Here, a new model is introduced to more accurately simulate a spherical particle using a 2D approach. The physical configuration of the experiment using a spherical particle is shown in Fig. 3. The problem can be simplified and solved using a modified cylindrical coordinate, as shown in Fig. 4.

Fig. 3 Schematic of a spherical biomass particle in the experimental setup

Fig. 4 Axi-symmetry cylindrical coordinate

The 3D space can be approximated by the simulation area (shown in Fig. 3) rotated along the symmetry axis by 360 degrees using the cylindrical coordinate ( r ,θ , z ) (Fig. 4). As a result, the 3D problem can be approximated using a 2D approach based on the cylindrical coordinate,

11

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 33

since the variation along θ direction can be neglected. The energy equation Eq. (21) can be solved in cylindrical coordinate as Eq. (22).

∂ ( ρC pT ) + ∇ ⋅ ( ρ CPTu) = λ∇2T + Q ∂t

(21)

∂ ∂T 1 ∂T ∂T 1 ∂ ∂T 1 ∂ ∂T ∂ ∂T ( ρC pT ) + ( ρCP u)( + + )= (λ r ) + 2 (λ ) + (λ ) + Q , ∂t ∂r r ∂θ ∂z r ∂r ∂r r ∂θ ∂θ ∂z ∂z 0 ≤ θ ≤ 2π In the above equation,

(22)

1 ∂T 1 ∂ ∂T = 2 (λ ) = 0 , since the variation along θ direction is not r ∂θ r ∂θ ∂θ

considered. Thus, the energy equation can be simplified. ∂ ∂T ∂T ∂ 2T λ ∂T ∂ 2T ( ρ C pT ) + ( ρ C P u)( + )=λ 2 + +λ 2 +Q ∂t ∂r ∂z ∂r r ∂r ∂z

(23)

This above equation is compared to the energy equation on the Cartesian coordinates. ∂ ∂T ∂T ∂ 2T ∂ 2T ( ρ C pT ) + ( ρ C P u)( + )=λ 2 +λ 2 +Q ∂t ∂r ∂z ∂r ∂z

Only the additional term

λ ∂T r ∂r

(24)

needs to be implemented into the original energy equation,

which can be solved explicitly by using a central differencing scheme. This new term can be viewed as the effect of heat transfer from the “third direction.” Thus, a 3D problem involving a spherical particle can be solved by using the modified energy equation (Eq. (23)).

3

Validation To examine the predictive capability of the present numerical method, two different

studies15,33 were chosen for comparison. In this section, the material properties, simulation conditions, and results are described.

12

ACS Paragon Plus Environment

Page 13 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

3.1

Case 1: Study by Ciesielski et al.33 Ciesielski et al.33 computationally studied the intra-particle heat and mass transfer of

biomass particles using detailed 3D finite element simulations to resolve the microstructure in the particle. They found that the numerical results considering realistic morphology deviated from those using conventional spherical models used in most numerical studies. In their paper, the cylindrical particle with an elliptical cross section (Fig. 5) was situated in an environment with initial temperature 25 °C and pressure 1 atm. A wall temperature of 500 °C was applied to the boundaries of the simulation domain to heat the gas and particle. The particle was pine with the following properties: density 540 kg/m3, thermal conductivity 0.12 W/(m K), and specific heat CP = 0.1031+ 0.003867T kJ/(kg K).

Fig. 5 Schematic of the simulation domain

13

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a)

Page 14 of 33

(b)

Fig. 6 Comparisons of the present and previous results33 for a pine particle heated by 500 °C walls: (a) center temperature; (b) volume-averaged temperature

In the present study, 2D simulations were conducted with the domain shown in Fig. 5. Fig. 6 shows the comparisons of results obtained by Ciesielski et al.33 and the present study. It can be seen from Fig. 6(a) that, the present 2D simulation results using LBM are very close to the 3D detailed finite element simulation data on center temperature. As shown in Fig. 6(b), the volumeaveraged temperature obtained by the present study was slightly lower than that of Ciesielski et al.33. One possible reason is the discrepancy between 2D and 3D simulations in that the 2D case neglected the effects of heat transfer in the axial direction. Despite the deviation, the trend and final temperatures are well predicted, indicating that the present method was able to capture the heat transfer characteristics of biomass particles during pyrolysis. On the other hand, the present 2D simulation is more computationally efficient. 3.2

Case 2: Study by Park et al.15 Park et al.15 studied the intra-particle mass and heat transfer processes during wood

pyrolysis. In their experiments, wood spheres with 25.4 mm diameter were pyrolyzed at temperature ranging from 638 K to 879 K. The temperature and mass losses of the biomass

14

ACS Paragon Plus Environment

Page 15 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

particle were recorded in the experiment. The physical properties of the biomass particle are shown in Table 2. Initial porosity of biomass particle was 0.4. Table 2 Physical properties of the biomass particle13,15

Property Density of wood,

Value

ρw

630 (kg/m3)

Specific heat capacity of wood, CPw

1500+1.0T (J/kg K)

Specific heat capacity of char, CPc

420+2.09T+6.85×10-4T2 (J/kg K)

Specific heat capacity of tar, CPt

−100+4.4T−1.57×10-3T2 (J/kg K)

Specific heat capacity of syngas, CPsyn

770+0.629T−1.91×10-4T2 (J/kg K)

Thermal conductivity of wood, λw

0.20487 (W/m K)

Thermal conductivity of char, λc

0.0937 (W/m K)

λg

0.0258 (W/m K)

Thermal conductivity of gas, Viscosity of gas, µ

3.0×10-5 (kg/m s)

Universal gas constant, R

8.314 (J/ mol K)

Stefan-Boltzmann constant Pore diameter d

σ

5.67×10-8 (W/m2 K4) 5×10-5(1- η )+1×10-4 η (m)

In order to consider the compositional change and radiative heat transfer during the conversion of the biomass particle, the effective thermal conductivity was used, as proposed by Park et al.15.

λ eff = (1 − η ) λ w + ηλ c + ελ g + 13.5σ T 3 d

(25)

σ and d are the Stefan-Boltzmann constant and pore diameter, respectively. The specific heat was calculated as a function of temperature as shown in Table 2.

15

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 33

In general, the temperature for pyrolysis is around 500 °C. Thus, the 783 K case from Park’s experiment15 was chosen for comparison (i.e., reactor wall temperature 783K and inlet gas temperature 720 K). The predicted solid mass fraction and center temperature were compared with the experimental data in Fig. 7. The solid mass fraction was defined as the mass ratio of the solid phase (sum of unreacted biomass, intermediate solid, and char) to the virgin biomass. Good levels of agreement can be seen for the spherical particle (using the improved 2D model described above). The predicted solid mass fraction was slightly higher than the experimental data, while the trend and final mass ratio were well predicted within 3.25% error. The calculated peak temperature is slightly over-predicted within 1.58% error but occurs approximately 30 s later than the experiment. However, the shapes of the curves are very similar. Considering the possible uncertainties in both experiments and simulations, as well as the complexity of the problem, the agreement is relatively good. For the results of cylindrical particle (using the original LBM model), the predicted center temperature increased more slowly, and the conversion process took longer time to complete.

Fig. 7 Comparison between measured and predicted solid mass fraction and center temperature (783 K reactor wall, 720 K inlet gas)

16

ACS Paragon Plus Environment

Page 17 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Fig. 8 Predicted pyrolysis product mass fractions (lines), compared with the numerical results at the final stage15 (symbols)

The temporal evolutions of the pyrolysis product yields are shown in Fig. 8. The final product yields predicted by the present model were found to be very close to the measurements15. As the particle temperature increased, the reaction rates were accelerated and the product yields increased rapidly at the beginning. The intermediate solid yield increased and then decreased, as the reaction mechanism suggested. It can be seen that at 380 s, the tar and syngas yields reached the maximum, while the char yield kept increasing until around 400 s. After that, the product yields remained unchanged as pyrolysis had completed and only char existed in the solid phase.

17

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a)

(b)

(c)

(d)

(e)

(f) Fig. 9 Snapshots of the pyrolysis process at 100 s, 200 s, 300 s and 400 s (a) temperature (b) biomass density (c) tar density (d) syngas density (e) intermediate solid density (f) char density

18

ACS Paragon Plus Environment

Page 18 of 33

Page 19 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Snapshots of important properties during the conversion process are illustrated in Fig. 9. Fig. 9(a) shows that the temperature in the particle was not uniform during the process (from 100 s to 400 s). The biomass particle was initially at a lower temperature and was heated up rapidly. This is also partly because the inlet gas temperature was lower than the reactor wall temperature, which would affect the heat exchange between the gas and particle. The influence of the nonuniform temperature on biomass pyrolysis are shown in Fig. 9(b)-7(f). From Fig. 9(b), the biomass density was not uniform in the particle because of the difference in the reaction rate caused by temperature gradients. As time progressed from 100 s to 400 s, biomass was consumed gradually and the products were generated. At 400 s, the pyrolysis process was nearly completed, thus the biomass density approached zero. Fig. 9(c) and 7(d) show the spatial distributions of tar and syngas density, respectively. Tar and syngas were produced in the particle and diffused into the gas flow and finally left the reactor. The evolutions of the intermediate solid and char are presented in Fig. 9(e) and 9(f). It can be seen that the intermediate solid was produced and consumed subsequently, consistent with the results shown in Fig. 8. The change in char density was also anisotropic, as shown in Fig. 9(f).

4 4.1

Parametric study Effect of particle size It has been found that the size of biomass particles will affect the thermochemical

conversion process as well as the product yields11,14. Here, a parametric study was conducted for both spherical and cylindrical particles with diameters of 0.5, 2.5, 10, and 50 mm. The inlet gas temperature was 720 K and reactor wall temperature was 783 K. The final product yields with respect to different particle sizes are shown in Fig. 10. The results indicate that the final tar and syngas yields decreased while char yield increased as the diameter of the particle increases for 19

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 33

both spherical and cylindrical particles, consistent with the finding in the literature14,17. This trend can be explained by the fact that the char yield is favored at low pyrolysis temperature while tar and syngas yields increase at high temperature. Under the same gas and reactor temperatures, increasing the particle size will result in a slow heat transfer into the particle. Hence, more char will be produced. In addition to the final product yields, the conversion time is also a key factor in biomass fast pyrolysis. Here, the conversion time was defined as the time when pyrolysis reached 95% conversion extent. The effect of particle size on the conversion time is illustrated in Fig. 11. It took a longer time to heat and convert the particle with a larger size as the conversion time increased significantly when the particle size increased from 0.5 mm to 25 mm. The impact of particle size on the conversion time is more significant for the cylindrical particle. It is worth noting that, for the 0.5-mm particle, it is very difficult to observe the conversion process in experiments. Thus, the present numerical method can be an effective tool to characterize the effects of particle size.

Fig. 10 Final product yields for particles with different sizes (783 K reactor wall, 720 K inlet gas)

20

ACS Paragon Plus Environment

Page 21 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Fig. 11 Conversion time for particles with different sizes (783 K reactor wall, 720 K inlet gas)

Generally, for large particles, it is important to characterize the intra-particle transport because of the highly non-uniform temperature. It is informative to determine the threshold particle size below which the particle can be treated as homogeneous and the effect of intraparticle gradients can be neglected. The study here is to investigate the effect of particle size on intra-particle gradients by assessing the rate of heat transfer and chemical reactions inside the particle. Fig. 12 shows the transient profiles of the solid mass fractions and center temperatures of particles (both sphere and cylinder) with diameters ranging from 0.1 to 1 mm. It can be seen that, for a spherical particle with diameter 0.1 and 0.2 mm, the center temperature increased very quickly and reached the final steady-state temperature in a very short time. The histories of the solid mass fractions of these two cases were nearly identical. This indicates that when the conversion process started, the temperature in the particle can be considered homogenous. Under such conditions, the temperature gradients can be neglected and the pyrolysis process can be considered kinetics-controlled.

21

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 33

(a)

(b)

(c)

(d)

(e) Fig. 12 Center temperature and solid mass fraction for particle size of (a) 0.1 mm (b) 0.2 mm (c) 0.5 mm (d) 0.75 mm (e) 1 mm

When the particle diameter increased above 0.5 mm, the solid mass fraction did not decrease noticeably until the center temperature reached the maximum. Prior to reaching the maximum temperature, approximately 20% of the biomass has been converted. During this

22

ACS Paragon Plus Environment

Page 23 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

period of time, the conversion occurred simultaneously with the heat transfer in the particle. Under these conditions, the temperature gradients cannot be neglected. Note that the residence of a biomass particle in a typical fast pyrolysis reactor ranges from a fraction of a second to a few seconds. The preparation (e.g., grinding) of biomass feedstock will need to be consistent with the operating conditions of the reactor (e.g., residence time, gas temperature) to ensure favorable product yields. The results of the cylindrical and spherical particles are very similar. Comparisons of Fig. 12 and Fig. 7 show that the difference between the results of spherical and cylindrical particles increased with the increase in particle size. For particle size less than 0.2 mm, the difference was very small, under which conditions the spherical particle can be simulated by using a 2D model. 4.2

Effect of particle aspect ratio For cylindrical particles, the aspect ratio is a key factor impacting the conversion process. In

order to investigate the effect of particle aspect ratio on the conversion time and pyrolysis outcome, cylindrical particles with different aspect ratios were simulated. The particle aspect ratio is defined by a:b as shown in Fig. 13. For example, the aspect ratio of a circular crosssection cylinder is 1:1. Note that the major radius and minor radius were adjusted such that the cross-sectional area of the elliptic particle was the same as that of the circular particle. In addition, the flow area between the particle and the reactor wall was also kept the same.

23

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 33

Fig. 13 Definition of particle aspect ratio

(a)

(b)

(c)

(d)

Fig. 14 Effects of particle aspect ratio on (a) conversion time (b) tar yield (c) syngas yield (d) char yield

24

ACS Paragon Plus Environment

Page 25 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Fig. 15 Comparisons of center temperatures for particles with different particle aspect ratios under the same heating conditions

The conversion time and final product yields for particles with different aspect ratios are shown in Fig. 14. It can be seen that, the conversion time increased with the increase in particle aspect ratio initially and then decreased when the particle aspect ratio further increased over 1:1. As the particle aspect ratio increased, tar and syngas yields decreased at first and then increased. The char yield had an opposite trend to syngas and tar. It is interesting to note that the noncircular particle has a lower conversion time than the circular particle. The reason is that the temperature in a non-circular particle increased more quickly than that in a circular particle. The evolutions of center temperature for different particle aspect ratios are shown in Fig. 15. It can be seen that the center temperature of a non-circular particle increased faster than the circular particle. Compared to the circular shape, an elliptic particle is flatter and thermally thinner; therefore, heat can penetrate into the particle more easily. Additionally, the perimeter of an ellipse is larger than that of a circle with the same area, which is beneficial for the convective

25

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 33

heat transfer from the surrounding gas. This effect is more significant for flatter particles (i.e. 1:4 and 4:1). 4.3

Effect of inlet gas temperature The heating to biomass particles in a reactor is provided by the inlet gas and the reactor wall,

which is maintained at a high temperature. The effect of inlet gas temperature (Tin) on biomass fast pyrolysis was investigated here. The conversion time and final product yields for different Tin are shown in Fig. 16.

(a)

(b)

(c)

(d)

Fig. 16 Effects of inlet gas temperature on (a) conversion time (b) tar yield (c) syngas yield (d) char yield

It can be seen that for both spherical and cylindrical particles, the conversion time decreased linearly as the inlet gas temperature increased from 700 K to 783 K. When the inlet gas

26

ACS Paragon Plus Environment

Page 27 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

temperature further increased above 783K, the conversion time remained nearly the same. As Tin increased, tar and syngas yields increased at first and then remained unchanged for 783 K and above. The char yield had an opposite trend to syngas and tar. A high Tin will result in a high particle temperature and accelerate the conversion process to produce more tar and syngas with less char. But a further increase in Tin only have a limited effect on the conversion process. 4.4

Effect of reactor wall temperature The effect of reactor wall temperature (Twall) on biomass fast pyrolysis was also investigated.

The conversion time and final product yields at different Twall are shown in Fig. 17. It can be seen that the conversion time decreased rapidly as the reactor wall temperature increased from 700 K to 900 K for both spherical and cylindrical particles.

(a)

(b)

(c)

(d)

Fig. 17 Effects of reactor wall temperature on (a) conversion time (b) tar yield (c) syngas yield (d) char yield

27

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 33

On the final product yields, as Twall increased, tar and syngas yields increased while the char yield decreased. With a high Twall, the particle temperature also increased, producing more tar and syngas but less char. When compared with the effect of inlet gas temperature, it can be found that the heat flux from the reactor wall was more dominant in the fast pyrolysis of biomass particles under the conditions studied. Note that this work is a detailed numerical study at the particle scale. It is hoped that the present numerical method will be further developed into a high-fidelity simulation tool to predict the performance of a large-scale reactor and accelerate the widespread deployment of biomass fast pyrolysis technology.

5

Conclusions The fast pyrolysis process of a single stationary biomass particle was simulated using a

detailed numerical method that resolved the flow field at the particle level. The Lattice Boltzmann method was used to predict the intra-particle transport of mass and heat coupled with the surrounding gas flow. A modified 2D model was used to simulate the conversion of spherical particle. The numerical results agreed with experimental data in the evolutions of the center temperature and solid mass fraction. The final product yields were also predicted well. Parametric studies were conducted to examine the conversion time and final product yields at different conditions. The baseline inlet gas temperature was 720 K and reactor wall temperature was 783 K. The time for complete pyrolysis increased as the particle size increased. For a small particle, the temperature in the particle increased quickly, resulting in more tar and syngas at the completion of the pyrolysis process. For particle diameter equal to or less than 0.2 mm, the temperature gradients in the particle was negligible and the pyrolysis process could be

28

ACS Paragon Plus Environment

Page 29 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

considered homogenous. For particle size less than 1 mm, uniform temperature in the particle was achieved within one second under the conditions studied. When the particle diameter equal to or smaller than 0.2 mm, the results of cylindrical and spherical particles are very similar. This indicates that a spherical particle can be treated as a cross section on the axial direction of a cylindrical particle if the diameter is small. The effect of particle aspect ratio was also examined. The simulation results show that, the non-circular particle has higher center temperature and shorter conversion time than the circular particle. More tar and syngas are produced and less char is generated for non-circular particle. This is due to the larger surface area and the shorter distance from the center to the surface, compared to a circular particle. Simulations also showed that tar and syngas yields increased and the char yield decreased as the inlet gas temperature or reactor wall temperature increased. The impact of inlet gas temperature on the conversion process was not linear. Increasing the inlet gas temperature beyond 783 K did not affect the conversion time or the final product yields. On the other hand, increasing the reactor wall temperature resulted in a monotonic increase in tar and syngas yields even beyond 783 K. The effects of the reactor wall temperature appeared to be more significant in providing heat to the fast pyrolysis of biomass particles than the inlet gas under the conditions studied.

Acknowledgements This work was supported by the National Science Foundation under the Grant Number EPS-1101284.

29

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 33

References 1. Brown, R. C.; Brown, T. R., Biorenewable resources: engineering new products from agriculture. John Wiley & Sons: 2013. 2. Bridgwater, A. V., Review of fast pyrolysis of biomass and product upgrading. Biomass and bioenergy 2012, 38, 68-94. 3. Goyal, H.; Seal, D.; Saxena, R., Bio-fuels from thermochemical conversion of renewable resources: a review. Renewable and sustainable energy reviews 2008, 12, (2), 504-517. 4. Czernik, S.; Bridgwater, A., Overview of applications of biomass fast pyrolysis oil. Energy & Fuels 2004, 18, (2), 590-598. 5. Hossain, A.; Davies, P., Pyrolysis liquids and gases as alternative fuels in internal combustion engines–A review. Renewable and sustainable energy reviews 2013, 21, 165-189. 6. Mettler, M. S.; Vlachos, D. G.; Dauenhauer, P. J., Top ten fundamental challenges of biomass pyrolysis for biofuels. Energy & Environmental Science 2012, 5, (7), 7797-7809. 7. Di Blasi, C.; Branca, C.; Masotta, F.; De Biase, E., Experimental analysis of reaction heat effects during beech wood pyrolysis. Energy & Fuels 2013, 27, (5), 2665-2674. 8. Dupont, C.; Commandre, J.-M.; Gauthier, P.; Boissonnet, G.; Salvador, S.; Schweich, D., Biomass pyrolysis experiments in an analytical entrained flow reactor between 1073K and 1273K. Fuel 2008, 87, (7), 1155-1164. 9. Bharadwaj, A.; Baxter, L. L.; Robinson, A. L., Effects of intraparticle heat and mass transfer on biomass devolatilization: experimental results and model predictions. Energy & fuels 2004, 18, (4), 1021-1031. 10. Blondeau, J.; Jeanmart, H., Biomass pyrolysis at high temperatures: Prediction of gaseous species yields from an anisotropic particle. Biomass and bioenergy 2012, 41, 107-121. 11. Bates, R. B.; Ghoniem, A. F., Modeling kinetics-transport interactions during biomass torrefaction: The effects of temperature, particle size, and moisture content. Fuel 2014, 137, 216-229. 12. Di Blasi, C., Modeling chemical and physical processes of wood and biomass pyrolysis. Progress in Energy and Combustion Science 2008, 34, (1), 47-90. 13. Grønli, M. G.; Melaaen, M. C., Mathematical model for wood pyrolysis comparison of experimental measurements with model predictions. Energy & Fuels 2000, 14, (4), 791-800.

30

ACS Paragon Plus Environment

Page 31 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

14. Lu, H.; Ip, E.; Scott, J.; Foster, P.; Vickers, M.; Baxter, L. L., Effects of particle shape and size on devolatilization of biomass particle. Fuel 2010, 89, (5), 1156-1168. 15. Park, W. C.; Atreya, A.; Baum, H. R., Experimental and theoretical investigation of heat and mass transfer processes during wood pyrolysis. Combustion and Flame 2010, 157, (3), 481494. 16. Kung, H.-C., A mathematical model of wood pyrolysis. Combustion and flame 1972, 18, (2), 185-195. 17. Haseli, Y.; Van Oijen, J.; De Goey, L., Numerical study of the conversion time of single pyrolyzing biomass particles at high heating conditions. Chemical Engineering Journal 2011, 169, (1), 299-312. 18. Aramideh, S.; Xiong, Q.; Kong, S.-C.; Brown, R. C., Numerical simulation of biomass fast pyrolysis in an auger reactor. Fuel 2015, 156, 234-242. 19. Mellin, P.; Kantarelis, E.; Yang, W., Computational fluid dynamics modeling of biomass fast pyrolysis in a fluidized bed reactor, using a comprehensive chemistry scheme. Fuel 2014, 117, 704-715. 20. Xiong, Q.; Aramideh, S.; Kong, S.-C., Modeling effects of operating conditions on biomass fast pyrolysis in bubbling fluidized bed reactors. Energy & Fuels 2013, 27, (10), 5948-5956. 21. Xue, Q.; Dalluge, D.; Heindel, T.; Fox, R.; Brown, R., Experimental validation and CFD modeling study of biomass fast pyrolysis in fluidized-bed reactors. Fuel 2012, 97, 757-769. 22. Yu, X.; Hassan, M.; Ocone, R.; Makkawi, Y., A CFD study of biomass pyrolysis in a downer reactor equipped with a novel gas–solid separator-II thermochemical performance and products. Fuel Processing Technology 2015, 133, 51-63. 23. Xiong, Q.; Kong, S.-C.; Passalacqua, A., Development of a generalized numerical framework for simulating biomass fast pyrolysis in fluidized-bed reactors. Chemical Engineering Science 2013, 99, 305-313. 24. Xiong, Q.; Kong, S.-C., Modeling effects of interphase transport coefficients on biomass pyrolysis in fluidized beds. Powder Technology 2014, 262, 96-105. 25. Biswas, A. K.; Umeki, K., Simplification of devolatilization models for thermally-thick particles: Differences between wood logs and pellets. Chemical Engineering Journal 2015, 274, 181-191.

31

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 33

26. Galgano, A.; Di Blasi, C.; Horvat, A.; Sinai, Y., Experimental validation of a coupled solidand gas-phase model for combustion and gasification of wood logs. Energy & fuels 2006, 20, (5), 2223-2232. 27. Sand, U.; Sandberg, J.; Larfeldt, J.; Fdhila, R. B., Numerical prediction of the transport and pyrolysis in the interior and surrounding of dry and wet wood log. Applied Energy 2008, 85, (12), 1208-1224. 28. Chen, S.; Doolen, G. D., Lattice Boltzmann method for fluid flows. Annual review of fluid mechanics 1998, 30, (1), 329-364. 29. Succi, S., The lattice Boltzmann equation: for fluid dynamics and beyond. Oxford university press: 2001. 30. Machado, R., Numerical simulations of surface reaction in porous media with lattice Boltzmann. Chemical engineering science 2012, 69, (1), 628-643. 31. Li, X.; Cai, J.; Xin, F.; Huai, X.; Guo, J., Lattice Boltzmann simulation of endothermal catalytic reaction in catalyst porous media. Applied Thermal Engineering 2013, 50, (1), 1194-1200. 32. Feng, Y.; Han, K.; Owen, D., Coupled lattice Boltzmann method and discrete element modelling of particle transport in turbulent fluid flows: Computational issues. International Journal for Numerical Methods in Engineering 2007, 72, (9), 1111-1134. 33. Ciesielski, P. N.; Crowley, M. F.; Nimlos, M. R.; Sanders, A. W.; Wiggins, G. M.; Robichaud, D.; Donohoe, B. S.; Foust, T. D., Biomass particle models with realistic morphology and resolved microstructure for simulations of intraparticle transport phenomena. Energy & Fuels 2014, 29, (1), 242-254. 34. Larfeldt, J.; Leckner, B.; Melaaen, M. C., Modelling and measurements of heat transfer in charcoal from pyrolysis of large wood particles. Biomass and Bioenergy 2000, 18, (6), 507514. 35. Qian, Y.; d'Humières, D.; Lallemand, P., Lattice BGK models for Navier-Stokes equation. EPL (Europhysics Letters) 1992, 17, (6), 479. 36. He, X.; Chen, S.; Doolen, G. D., A novel thermal model for the lattice Boltzmann method in incompressible limit. Journal of Computational Physics 1998, 146, (1), 282-300. 37. Guo, Z.; Zhao, T., A lattice Boltzmann model for convection heat transfer in porous media. Numerical Heat Transfer, Part B 2005, 47, (2), 157-177.

32

ACS Paragon Plus Environment

Page 33 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

38. Seddiq, M.; Maerefat, M.; Mirzaei, M., Modeling of heat transfer at the fluid–solid interface by lattice Boltzmann method. International Journal of Thermal Sciences 2014, 75, 28-35. 39. Wang, J.; Wang, M.; Li, Z., A lattice Boltzmann algorithm for fluid–solid conjugate heat transfer. International journal of thermal sciences 2007, 46, (3), 228-234. 40. Zou, Q.; He, X., On pressure and velocity boundary conditions for the lattice Boltzmann BGK model. Physics of Fluids (1994-present) 1997, 9, (6), 1591-1598.

Nomenclature A CP D E k P Q R T t U

pre-exponential factor (s-1) specific heat capacity (J kg-1 K-1) mass diffusion coefficient (m2 s-1) activation energy (J mol-1) reaction rate (s-1) pressure (Pa) heat generation (W m-3) universal gas constant (J mol-1 K-1) temperature (K) time (s) gas velocity (m s-1)

Greek letters α thermal diffusivity (m2 s-1) ε porosity η conversion extent of pyrolysis ∆h heat of reaction (J kg-1) λ thermal conductivity (W m-1 K-1) µ viscosity (kg m-1 s-1) ν kinematic viscosity (m2 s-1)

ρ τ ω

density (kg m-3) relaxation time rate of production (kg m-3 s-1)

Subscripts ave average value c char, primary char generation reaction c2 secondary char generation reaction f fluid phase (tar, syngas) g gas i specie i, lattice direction i is intermediate solid j reaction j s solid phase syn syngas, primary syngas generation reaction syn2 secondary syngas generation reaction t tar, primary tar generation reaction T temperature w wood in inlet condition

33

ACS Paragon Plus Environment